If checkrad were changed to only output a 0 instead of the dump it currently produces when it can't connect to snmp on the NAS I think that could possibly correct the problem, I think that radiusd only expects a 1 character response, maybe it flips when it gets 7 lines of information?
Just an idea. Then again... checkrad might not be printing anything, could be just warnings from snmpget. I didn't look it to closely, I just know that it's working and is stable now - and I can't login twice. Adam Kristina Pfaff-Harris said: > On Fri, 21 Feb 2003, Alan DeKok wrote: > >> That's confusing as all get out. The code which runs checkrad >> tries >> to kill it if checkrad doesn't respond. But it sends *checkrad* a >> TERM signal, and doesn't send one to the RADIUS server. > > The wierd thing is, checkrad DOES respond -- it just responds with a > timeout, and if that weren't strange enough, checkrad appears to be > exiting normally, that is, it's not hanging or anything. > >> A short-term solution is to go to the bottom of src/main/session.c, >> and delete the 3 lines doing: >> >> kill(pid, SIGTERM); >> sleep(1); >> kill(pid, SIGKILL); > > Heh. Any ideas on where to start looking to track down a long-term > solution? :-) > > K. > > > - > List info/subscribe/unsubscribe? See > http://www.freeradius.org/list/users.html - List info/subscribe/unsubscribe? See http://www.freeradius.org/list/users.html
